[css-backgrounds] background-position definition doesn't reflect reality.

[emilio]

https://drafts.csswg.org/css-backgrounds/#background-position doesn't mention at all background-position-x or background-position-y, of which background-position is a shorthand of.

These properties should be defined. For example, Gecko accepts background-position-y: <keyword> <length>? (due to the legacy 3-value syntax) while Blink / WebKit don't.

cc @ericwilligers, you did some yak-shaving re. 3-value position syntax and such, do you know which syntax does background-position-* accept in Blink?

Is there any plan to avoid supporting the three-value positions in background-position as well?

[frivoal]

They are in level 4: background-position-x and background-position-y.

[fantasai]

@emilio Can you a) use more specific issue titles (where practical, this is practical) b) don't file multiple issues into one issue? :/

Is there any plan to avoid supporting the three-value positions in background-position as well?

Not really, unless someone proves there's no Web-compat issue with removing it. Seems unlikely since it's been around for a very long time now.

[emilio]

Yeah, sorry. Will try, thanks @fantasai :)

They are in level 4: background-position-x and background-position-y.

Looks like there's no interoperability here, but given they're in level 4 the issue as filed is invalid. Do we want to track the lack of interop here? Or should we close this?

[ewilligers]

Is there any plan to avoid supporting the three-value positions in background-position as well?

No - three-value position usage in background-position is about 5%.

do you know which syntax does background-position-* accept in Blink?

All browsers support the Level 3 spec grammar for background-position - see wpt.fyi.

Tests should be added for the Level 4 grammar and for background-position-x / background-position-y.